Tezzbuzz Desk: While Artificial Intelligence i.e. AI is bringing rapid changes in the world of technology, its use is also increasing for cyber crimes. According to Taiwanese cybersecurity research firm TeamT5, China-linked hacking groups have more than doubled the number of their cyberattacks after they started using DeepSeek and other open-source AI models. According to the report, hackers are creating malicious software and attack codes with the help of AI to complete common and repetitive tasks faster.
According to research, DeepSeek is becoming especially popular among Chinese hackers. The main reasons for this are said to be its low cost, good capacity, facility of customization and relatively less restrictions related to cyber security. According to Charles Lee, chief analyst at TeamT5, there is also demand for Western AI models, but they have more stringent security guardrails. In such a situation, hackers have to make more efforts to misuse them. AI is being used not just in one phase of cyber attacks, but at many levels. TeamT5 has reported finding scripts and logs that show that hacking groups linked to China are using AI to gather information about targets, find system vulnerabilities and create code to be used against them. However, the researchers also clarified that it was not possible to accurately identify the AI model used in every cyber attack.
The report also revealed examples of several hacking groups. According to TeamT5, a group named Grimfengxi allegedly created the exploit code with the help of DeepSeek. A group named Huapi targeted the email system of a Taiwanese company using a Chinese AI model, which researchers believed to be DeepSeek. A group named Teleboyi allegedly collected about 1,000 IP addresses on the Internet with the help of AI and mapped the domain of a company. Interestingly, it is not only Chinese AI models that are being used in cyber attacks. According to the report, evidence of use of ChatGPT and Anthropic's Claude Code has also been found in Chinese hacking activities. In one case, hackers allegedly tried to create software with the help of ChatGPT to decrypt the Signal database obtained from an employee's compromised computer. TeamT5 has said that a group named Slime22 has used Claude Code after entering the system of a Taiwanese tech company.
This entire incident has created new concerns regarding cyber security. With the help of AI, experienced hackers can work on more targets in less time and increase the scale of cyber attacks. However, this does not mean that the AI is successfully carrying out every attack completely on its own. Some recent research has also highlighted the role of human intervention in AI-based attack attempts. = Overall, the use of DeepSeek and other AI models in cyberattacks shows that AI is no longer just a tool for productivity and technological advancement. To prevent its misuse, the need for strong security guardrails, surveillance and cyber defense systems has increased more than ever.